INTERNATIONAL

overty Centre P Country Study number 9 September, 2007

The International Poverty Centre is jointly supported by the Brazilian Institute for Applied Economic Research (IPEA) and the Bureau for Development Policy, United Nations Development Programme, New York.

COUNTRY STUDY CASH TRANSFER RESEARCH PROGRAMME

THE CHALLENGES OF ’S CONDITIONAL CASH TRANSFER PROGRAMME, RED SOLIDARIA

Tatiana Feitosa de Britto Visiting Researcher, International Poverty Centre

The views expressed in IPC publications are those of the authors and do not necessarily reflect the views of IPC, IPEA or UNDP. Copyright© 2007

International Poverty Centre United Nations Development Programme

International Poverty Centre SBS – Ed. BNDES,10o andar 70076 900 Brasilia DF Brazil [email protected] www.undp-povertycentre.org Telephone +55 61 2105 5000 Fax +55 61 2105 5001

Rights and Permissions

All rights reserved.

The text and data in this publication may be reproduced as long as the source is cited. Reproductions for commercial purposes are forbidden.

The International Poverty Centre disseminates these Country Studies of the Cash Transfer Research Programme to encourage the exchange of ideas about development issues. These studies are signed by the authors and should be cited accordingly. The findings, interpretations, and conclusions that they express are those of the authors. They do not necessarily represent the views of the International Poverty Centre, IPEA or the United Nations Development Programme, its Administrator, Directors, or the countries they represent.

Country Studies of the Cash Transfer Research Programme are available online at http://www.undp- povertycentre.org and subscriptions can be requested by email to [email protected] COUNTRY STUDY PUBLISHED BY THE INTERNATIONAL POVERTY CENTRE

THE CHALLENGES OF EL SALV AD OR’S COND ITIONAL CASH TRANSFER PROGRAM M E, RED SOLIDARIA∗ Tatiana Feitosa de Britto∗∗

CASH TRANSFER RESEARCH PROGRAM M E

This study is part of a much larger research programme on cash transfers in both Latin America and sub-Saharan Africa, supported by the Department for International Development, GTZ and UNDP.

ABSTRACT

In the context of the increasing prominence of conditional cash transfers (CCTs) in the development agenda of many developing countries, this Country Study provides an analytic overview of the challenges faced by El Salvador’s CCT programme, Red Solidaria, (Solidarity Netw ork). The purpose is to generate an information base for comparative studies on the prospects and potential difficulties of implementing CCTs in country settings different from those of the pioneer programmes, such as in Brazil and Mexico. The study describes Red Solidaria´s origins and components and discusses major aspects of its design and implementation. A particular emphasis is placed on the programme’s co-responsibilities, exit rules and targeting strategy. The study also covers the topics of institutional structures, intersectoral coordination and political support for such programmes. The conclusion is that Red Solidaria is an informative example of how a small country w ith limited resources can successfully set up a complex CCT programme. Still, the study notes that there are pending issues and remaining challenges for the programme. These relate, in particular, to strengthening mechanisms of local participation; coordinating the CCTs w ith other dimensions of Red Solidaria, such as productive projects; lengthening the duration of benefits for meeting human-capital objectives; clarifying eligibility requirements and how changes in family conditions can affect such requirements; and distinguishing conditionalities from ordinary programme co-responsibilities. An issue of overriding importance is to develop a broader long- term social protection strategy for El Salvador, w ith w hich CCTs w ould be integrated instead of being regarded as a stand-along programme.

∗ I would like to acknowledge the useful remarks received from Fabio Veras of the International Poverty Centre, who was the internal peer reviewer of this study, and Matthias Rompel and Nicola Wiebe of GTZ, who were the external peer reviewer. I also thank the Red Solidaria and FISDL staff for logistical support and collaboration during the research fieldwork in El Salvador. Thanks are also extended to Terry McKinley, Acting Director of IPC, who helped edit the final version of this study. ∗∗ Visiting Researcher, International Poverty Centre.

2 Country Study nº 9

1 INTROD U CTION

Conditional cash transfers (CCTs) have been frequently highlighted as an innovative tool of social protection. They combine the short-term objective of alleviating poverty w ith incentives for families to increase their investments in human capital so as to break intergenerational poverty traps. Since their first introduction in large, middle-income countries, such as Mexico and Brazil in the mid-1990s, they have become popular across Latin America, harnessing substantial support from multilateral agencies and donors. This Country Study discusses one of the most recent CCTs implemented in the region, Red Solidaria, initiated by the government of El Salvador in 2005. In the context of increasing CCT prominence in the region’s development agenda, the replicability and sustainability of these programmes in country settings w here financial and operational framew orks are w eaker is an issue of major concern. Thus, our purpose is to provide a descriptive and analytic overview of Red Solidaria that can generate inputs for comparative studies on the prospects and potential difficulties faced by CCTs in countries w ith settings different from those in the CCT pioneers. This study is based on data from three main sources: 1) a review of conceptual and operational documents and reports on Red Solidaria’s design and implementation; 2) semi- structured interview s w ith programme managers and stakeholders at the national, local and international levels; and 3) on-site observation of implementation processes carried out in three municipalities targeted by the programme. A list of interview s and field visits that have informed this study is presented in the Appendix. In Section 2 w e present an overview of Red Solidaria’s components and origins. Section 3 discusses key design and implementation aspects. Section 4 focuses on the programme’s co-responsibilities w hile Section 5 introduces the issue of exit rules or beneficiary graduation. Section 6 deals w ith targeting criteria and mechanisms. Section 7 discusses institutional structures and implementing capacity. Section 8 analyses the Red’s intersectoral co-ordination. Section 9 addresses the issue of sustainability, taking into account financial costs and political support. Section 10 concludes by identifying pending issues and challenges faced by the programme.

2 RED SOLIDARIA’S COM PONENTS AND ORIGINS

Red Solidaria is the main government programme targeted at the poorest population of El Salvador. Besides the cash transfer, it comprises tw o other components related to improvements in the supply of social services and infrastructure and to increases in the productivity and diversification of the income sources of poor families (see Box 1). The Red’s designers trace its origins to the government plan for 2004-2009 of President Antonio Elias Saca, a candidate of the right-w ing Alianza Republicana N acionalista (ARENA). This plan outlined the creation of a social safety net for the country’s most vulnerable population.1 Interestingly, the government plan of the leftist Frente Farabundo M artí para la Liberación N acional (FMLN) —the other strong contesting party in the 2004 elections— mentioned under its educational priorities the establishment of cash transfers to combat school drop-outs, low rates of attendance in school and child labour. International Poverty Centre 3

BOX 1 Com ponents of Red Solidaria

Component 1 Component 2 Component 3 Family Solidarity Network Network of Basic Services Family Sustainability Network

Cash transfer targeted to families Supply-side programmes in Promotion of productive projects with pregnant women and children education (Effective Schools and micro-credit schemes in the under 15 years old who have not Network), health and nutrition targeted municipalities. finished 6th grade, conditional on (Extension of Health Services), and basic health care activities and improvements and rehabilitation of school attendance. It also includes basic and strategic infrastructure lifelong learning sessions for (drinking water, sanitation, beneficiary families. electricity and rural roads). Source: Red Solidaria, 2006.

The programme’s conceptual design took place betw een 2004 and the beginning of 2005 and the Red w as officially launched by a presidential decree in March 2005.2 This legal document lays the foundations of Red Solidaria, in terms of objectives, duration, assigned resources and institutional coordination. It states that the programme’s objective is to assist extremely poor families through short-term improvements in: (1) child and maternal health and nutrition; (2) basic education; and (3) drinking w ater, sanitation supply, electricity and roads to the poorest rural communities of the country. The programme’s components are described as tools to broaden the opportunities of the disadvantaged population in order to improve their economic and social condition. The programme’s duration coincides w ith the presidential term and the coverage goals are the 100 poorest municipalities, as identified by the national poverty map (presented in Appendix B). Resources for all three components are estimated to be approximately US$ 50 million per year. The Social Investment Fund for Local Development (FISDL) is appointed as the implementing agency of the Red, but the programme’s technical and political coordination is placed under the Technical Secretariat of the Presidency and a Directive Council involving several government organizations. The focus on human capital investments w as clearly highlighted in Red Solidaria from its inception. Thus, although immediate income protection rests at its core, the programme’s logical framew ork describes the Red’s overarching goal as the “reduction of the intergenerational transmission of poverty in El Salvador”, and this is linked to the achievement of the Millennium Development Goals (MDGs) by 2015.3 The purpose is the medium- and long-term improvement of the living conditions of extremely poor rural families through integrated interventions that give priority to the poorest municipalities. No income indicator is used to assess the achievement of this goal. Instead, the programme’s logical framew ork uses increases in enrolment, school attainment and immunizations, combined w ith decreases in child malnutrition and the prevalence of diarrhoea as its core indicators. This illustrates the importance given to human capital in advancing tow ards the Red’s overall goal. The rural emphasis of the programme is another of its distinctive features. The rationale for this emphasis is related to the poverty profile of El Salvador, w hich documents the 4 Country Study nº 9

disproportionately disadvantaged situation of the rural population, not only in terms of income poverty but also in terms of access to health, education, other basic services and infrastructure (W orld Bank, 2005; PNUD, 2003). Although poverty w as significantly reduced betw een 1991 and 2002, the conditions in rural areas have remained critical. According to W orld Bank estimates, the national poverty headcount fell from 64 per cent to 37 per cent, and extreme poverty fell from 31 per cent to 15 per cent during 1991-2003. But 50 per cent of rural Salvadorans still remained below the poverty line in 2002 and 24.5 per cent w ere extremely poor. In urban areas these figures w ere 28.5 per cent and nine per cent respectively.4 Similarly, w hile basic education enrolment rates w ere close to 90 per cent in urban areas, they w ere only 80 per cent in rural areas. Differentials in access to safe w ater, adequate sanitation and electricity w ere even more striking.5 The design of the Red’s CCT and infrastructure components follow s from the recommendations emerging from this poverty assessment and a previous social safety net assessment, also sponsored by the W orld Bank. The latter had identified more than 50 safety net-like programmes operating in the country, most of them small-scale or pilot projects, often uncoordinated and overlapping (W orld Bank, 2005). These reports highlighted the need for crafting an integrated social strategy focusing on: a) strengthening human capital, mostly through increasing education levels and providing access to basic health services to all; b) strengthening people’s access to markets and services by improving the supply of drinking w ater, sanitation services and roads in rural communities; and c) assisting and protecting the poorest and most vulnerable population, by means of a coherent, coordinated and w ell-targeted social safety net.

3 D ESIGN AND IM PLEM ENTATION

The cash transfers provided by Red Solidaria comprise a health stipend for families w ith pregnant w omen and children under 5 years old and an education stipend for families w ith children from 5 to 15 years old w ho have not completed 6th grade, the second cycle of basic education. Each stipend is w orth US$ 15 per month per family, but a family cap applies to those entitled to both, i.e., US$ 10 per month for each, totalling a maximum of US$ 20 per family. There are no variations in the amount of the transfers according to the size of the family or by the age, grade or gender of beneficiary children. The flat stipend per family might be grounded on a theoretical rationale related to economies of scale in household consumption. Additionally, since the stipend favours smaller families, it might indicate a concern regarding fertility incentives (Handa and Davis, 2006). The programme does not seem, how ever, to take into account direct and indirect costs involved in complying w ith its conditionalities—such as costs for school material and transportation and the opportunity costs of w omen and children involved in the programme. Nor does it take into account variable time and transportation costs incurred by recipients to collect the transfers. These depend on the distance of households from the place w here payment takes place, w ith greater disadvantages for the more remote (and often poorer) communities. International Poverty Centre 5

3.1 THE SIZE OF TRANSFERS The criteria that informed the benefit size of Red Solidaria are not documented. Interview s w ith programme designers indicated that the decision on benefit values took into account the value of rural minimum w ages, as w ell as the international experience w ith CCTs. Red’s transfers are supposed to represent around one quarter of the nominal minimum w age in agricultural activities, set at US$ 74.1 per month in 2004 (Funde, 2005). It should also be noted that CCTs should try to balance the priority of alleviating poverty in the short run w ith promoting the longer-term objective of investments in human capital—w hile, at the same time, avoiding the encouragement of dependency on transfers (Cohen and Franco, 2006). Nonetheless, the extent to w hich most CCTs have actually taken such calculations of values into account in their design is not clear, as budget constraints and political considerations related to setting the maximum number of beneficiaries are also intervening factors. In any case, Handa and Davis (2006) argue that an international rule of thumb is that, in order to be meaningful for beneficiaries, a poverty-related cash transfer should represent 20-40 per cent of the per capita poverty line. Considering El Salvador’s official poverty line for rural areas, Red’s health and education stipends w ould lie w ithin that range (37 per cent), w hile the combined stipends w ould even exceed it (50 per cent).6 As w ith most CCTs implemented elsew here, Red Solidaria addresses its transfers primarily to the mother or another female family member w ho is in charge of children’s care.7 But the programme also includes a co-responsible beneficiary, usually the partner of the main female beneficiary or another person appointed by the family, w ho can w ithdraw the transfers in case the main beneficiary does not have her ow n identity card (a requirement to actually collect the stipend) or cannot be there in person during payment events. In order to minimize the occurrence of beneficiaries w ithout formal identification, and to increase their access to this basic right of citizenship, the programme also includes plans to coordinate efforts w ith the government agency responsible for the provision of the national identity card.

3.2 PAYMENT EVENTS The implementation of Red Solidaria, in particular its CCT component, has been a phased one. The first transfers took place in October 2005, w ith 15 municipalities included in the programme in that year, totalling 13,278 beneficiary families. In 2006, 17 additional municipalities and 10,828 families w ere included. This completed coverage of the first group of 32 municipalities characterized by very high extreme poverty, determined by the poverty map. In 2007, the programme intends to reach another 15 municipalities, covering around 45,000 families in total. By 2009, the goal of the programme is to reach 100,000 families in the 100 targeted municipalities. Transfers are delivered bi-monthly in payment events that usually take place in the dow ntow n area of the municipality and might occupy the w hole morning, depending on the number of beneficiaries in each specific municipality. The operation of the transfers is contracted out to a banking institution, w hich organizes the logistics of the payments, including the transport of the money, security and cashiers. In contrast to the experience of other countries that rely on a fairly developed and w idespread banking netw ork, w hich allow s an automated scheme for the delivery of CCTs through money teller machines and magnetic 6 Country Study nº 9

cards, Red’s transfers are handled manually and involve the displacement of considerable amounts of money and personnel across the country. The Red’s local coordination has an important role in the organization of payment events. An entire payment schedule is set up and beneficiaries are lined up according to this schedule and their exact placement in the transfer spreadsheet. This aims at rationalizing the process, shortening the time beneficiaries have to queue to receive their transfers. Nonetheless, there can be significant costs in terms of time and transportation for beneficiaries to attend the payment event. This strengthens the case for maintaining the Red’s bi-monthly delivery scheme.8

4 CO-RESPONSIBILITIES

As a CCT, Red Solidaria involves conditionalities that are usually associated w ith such transfers. These conditionalities are promoted as ‘co-responsibilities’, a term that implies a stronger role of beneficiaries in promoting their ow n w ell-being. Molineux (2006) points out that, together w ith the principles of participation and empow erment, the idea of co-responsibility or self- help is central to the so-called New Poverty Agenda that has come to dominate development discourse in the 1990s. In Latin America, in general, these principles have found fertile ground, as the region has undergone public sector reforms originating from fiscal crises and w idespread distrust of the State. Britto (2005), as w ell as Handa and Davis (2006), stresses political economy issues associated w ith the introduction of conditionalities in the cash transfer model that has been disseminated in the region. These have required the poor to ‘w ork’ for their money, focusing on promoting children’s human capital and supposedly empow ering beneficiaries to do w ithout state support in the future. The assumption is that such features w ould make these programmes more acceptable to taxpayers and voters. How ever, the idea of co-responsibility involves the notion of reciprocity, namely, the notion that obligations apply not only to recipients but also to the programme and service providers. In the case of Red Solidaria, co-responsibilities are detailed in a family agreement (convenio) that has to be signed by the main beneficiary and the family co-responsible if, once selected, the family agrees to take part in the programme. In accordance w ith this reciprocal notion, the agreement includes co-responsibilities for the beneficiary family as w ell as the government of El Salvador (see Box 2). The agreement also includes an explicit provision that the administration of the cash provided by the programme w ill be placed in the hands of the main beneficiary and an oath that all the information provided by the family for their inclusion in the programme is true. If the information is proved false, the oath w ould authorize immediate w ithdraw al of the family from the programme. On the government side, the agreement expresses the conditions under w hich the transfer shall be suspended. In the case of the education stipend, suspension occurs if any beneficiary children have at least four unjustified absences from school in a month.9 In the case of the health stipend, suspension occurs if the family a) does not attend a scheduled health check-up and does not comply w ith the immunizations protocol once or b) does not either attend the health check-up or comply w ith the immunizations protocol tw ice.10 International Poverty Centre 7

BOX 2 Red Solidaria’s co-responsibilities

For the beneficiary families For the government of El Salvador

Through the Ministry of Education, provide basic Ensure school enrolment and attendance of their education up to 6th grade to beneficiary children 5-14 year olds, from pre-school to 6th grade. between the ages of 5 and 14 years. Register the family in health programmes, attend Through the Ministry of Health, provide basic the child and maternal health check-ups and health services in health units or other institutions ensure compliance with the basic child and subcontracted for such purposes. maternal health protocols and immunizations. Attend the family training sessions offered by Promote lifelong learning sessions for beneficiary Red Solidaria. families. Deliver the cash transfer, conditional on school enrolment and regular attendance of the Use the transfers provided by Red Solidaria on beneficiary children (education stipend) and on the food consumption. periodical child and maternal health check-ups and immunizations for children under 5 years old. Source: Family agreement.

Thus, the actual conditionalities for the delivery of the transfers encompass the usual engagement in maternal and child preventive health activities and basic education enrolment and attendance for school-aged children that is instituted by most CCTs across Latin America. Nevertheless, the family agreement goes beyond those conditionalities w hen it sets up co-responsibilities related to the use of the transfers and the participation in lifelong learning sessions.

4.1 W HAT IS A CONDITIONALITY? In practtice, the commitment to use the cash transfers on food is not monitored by the programme. As a matter of fact, the information material handed to beneficiaries w hen they sign the agreement stresses that the cash transfers should be used for food and for ensuring that children dedicate more time to their studies. Programme managers at the national level not only seemed unfamiliar w ith this requirement, but also judged it as an undesirable and excessively restrictive one. At the grassroots level, the perceptions of most programme managers and implementers w ere that transfers should be used for the benefit of the children, be this for food consumption, school materials, clothing, medicines or other basic needs of the family. Furthermore, this is the message that seems to be passed on to beneficiaries, and is in line w ith the rationale of achieving efficiency and w ell-being gains through expenditure decisions made by the mothers themselves. In contrast, participation in the programme’s lifelong training sessions is indeed verified. These sessions are conducted by NGOs for family support and follow -up. These organizations are contracted by the Red’s implementing agency and are involved in many other important aspects of the programme’s implementation in the field, including monitoring and control of co-responsibilities. Even though not participating in these sessions w ould not entail a suspension of the cash transfers, attendance is officially monitored by the local NGO promoters. Most beneficiaries are unaw are that this co-responsibility actually does not have any effect on the cash benefit. 8 Country Study nº 9

In any case, training sessions are intended to be useful and empow ering for beneficiary families—especially for w omen, w ho usually have experienced very few opportunities for participating in organized activities outside the domestic sphere. Besides being an instrument for clarifying the Red’s functioning, the monthly w orkshops encompass capacity-building on issues such as children’s rights; gender equity; hygiene, nutrition and safe handling of food; domestic violence; the importance of formal education for children; reproductive health; child labour; community participation and organization; and MDGs. In informal conversations held w ith beneficiaries during field visits, training sessions w ere frequently noted as an important and useful aspect of the programme. Until recently, how ever, there had been no unified training protocol follow ed by the NGOs that w ork w ith beneficiary families. They w ould carry out training sessions and develop the suggested topics according to their ow n previous experience in the field. The Red’s implementing agency has now developed training modules for each of these themes, on the basis of a participatory methodology that can be applied to an audience w ith little formal education. It intends to distribute this material to all NGOs that take part in the programme. Capacity-building for the local promoters that conduct these training sessions is regarded by programme managers as an important side benefit of Red Solidaria, in the sense that most of the field personnel employed by the NGOs are young promoters recruited in their ow n community.

4.2 THE RATIONALE FOR CONDITIONALITIES The central importance attributed to the programme’s co-responsibilities emerges clearly in the discourse of programme managers. During interview s, stakeholders involved in managing Red Solidaria at national and local levels repeatedly expressed view s that low achievements in education and health by the Salvadoran poor w ere related not only to supply failures of these services, but to a cultural tendency of undervaluing investments in human capital and relying excessively on remittances from abroad. This w as the main justification given for the strong emphasis of the programme on the compliance w ith, and the verification of, its co-responsibilities, including those that w ere not automatically linked to the transfers, such as the participation in family training activities. Interestingly, the concern that cash transfers might generate beneficiary dependency seemed to be stronger w ith regard to increasing the co-responsibility of beneficiaries tow ards their ow n w ell-being than implementing strategies to graduate them from the programme. Issues of graduation have occupied the centre stage in the debate on many ‘older’ CCTs in Latin America. W ith regard to assessing results, Red Solidaria still has to undergo an external impact evaluation, in the form of a quasi-experiment planned to start in 2007 and include three subsequent rounds. For the time being, the available information on results comes from administrative records, w hich show considerable increases in school enrolments and child and maternal health check-ups. Comparing records from the first trimester of 2005 w ith those for the same period in 2006 in the first 15 municipalities targeted by the programme, w e see that enrolment increased by 23 per cent in pre-school, six per cent in grades 1-3 and nine per cent in grades 3-6. Children’s health check-ups had experienced, in turn, an increase of 47 per cent and maternal health check-ups, 42 per cent (Red Solidaria, 2007). International Poverty Centre 9

The impact evaluation might provide relevant information to assess these statistics and determine to w hat extent they are related to improvements in the supply of health and education services in the targeted municipalities or to the demand-side incentives created by the co-responsibilities of the programme.

5 EXIT RU LES

In line w ith the conceptual and operational guidelines of Red’s legal framew ork and operational manual, the family agreement explicitly states its three-year duration. That means that each family w ould be entitled to receive the transfers for these three years only. This clause seemed very clear to most beneficiaries and all programme managers and stakeholders interview ed during the research fieldw ork. W hat did not appear clear to them w as w hat w ould happen afterw ards: w ould the families simply stop receiving the transfer, regardless of the persistence of the eligibility conditions that had entitled them to be selected in the first place? Or w ould the programme have some kind of transition or phasing-out strategy that w ould eventually graduate families but avoid the interruption of the transfers in the middle of a school or immunization cycle? According to the Red’s operational manual, the exit strategy of the programme incorporates a re-evaluation of each family, w hich is undertaken after it receives the transfers for three years. This re-evaluation w ould take into account w hether families still had children under five year olds (to continue to be eligible for the health stipend), or children betw een five and 15 years old w ho had not finished primary education, i.e., 6th grade (w hich w ould entitle them to continue receiving the education stipend). Nevertheless, the manual stresses that the questionnaire and the criteria to be used in this re-evaluation process are subject to validation by the inter-institutional Directive Council of Red Solidaria and to the availability of financial resources from the government of El Salvador (Red Solidaria, 2006). So far, the issue of programme graduation or mechanisms for re-certification of eligibility does not seem to be an important one for El Salvador’s CCT agenda. On the one hand, this might be due to the novelty of the programme, w hich is less than a year and a half old and still has not had to deal w ith this potentially difficult issue. On the other hand, it might be related to the Red’s central objective: to improve human capital, rather than merely reducing current income poverty. In this regard, the acceptable ‘exit door’ from the programme w ould seem to be a demographic one instead of an ‘emancipation’ strategy that w ould hypothetically be able to lift beneficiaries out of poverty—and that w ould be contingent upon many other factors related to the labour market and the economic performance of the country, as w ell as upon a synergistic coordination among different social programmes.11 Most CCTs include, at least in principle, an explicit time limit after w hich families should be reassessed. The three-year period is a fairly common threshold.12 But the theoretical or empirical rationales—if any—that have governed the decisions w ith regard to these time limits are unclear (Cohen and Franco, 2006). They seem to be driven mostly by fiscal or political considerations, associated w ith fears that transfers might become financially unsustainable, be perceived as permanent entitlements and create dependency among beneficiaries. 10 Country Study nº 9

Some CCTs have, therefore, introduced transition rules, w hich include not only recertification strategies, but also declining stipends as a phasing-out mechanism. How ever, the variation across programme practices and discourses, as w ell as the difficulties associated w ith adopting a coherent and transparent exit strategy, have contributed to relegating graduation rules to the ‘unfinished agenda’ of CCTs in Latin America (de la Brière and Raw lings, 2006). A crucial issue in this respect is that graduation rules have to be shaped according to the programme’s central objective. From the perspective of human capital accumulation, w hich is the main objective stressed by Red Solidaria, short time limits—if strictly applied—might be counterproductive especially if they are not sufficient to cover the entire basic education or health cycle (Handa and Davis, 2006; Cohen and Franco, 2006; Soares and Ribas, forthcoming).

6 TARGETING CRITERIA AND M ECHANISM S

Red Solidaria uses tw o criteria for targeting beneficiaries. First, there is geographical targeting, w hich is based on a poverty map developed by the Latin-American Faculty of Social Sciences (FLACSO), under FISDL supervision and w ith support from the W orld Bank and Inter-American Development Bank. This poverty map has identified the municipalities that are to be included in the programme. Second, there is household targeting, w hich selects families that are eligible to be programme beneficiaries in the targeted municipalities. The poverty map is based on an expansion of the yearly Multiple Purpose Household Survey (Encuesta de Hogares de Propositos M ultiplos – EHPM), w hich has enabled the setting up of a unified database for the years 2001-2004 w ith data representative for the 262 Salvadoran municipalities. The map comprises tw o types of municipal classification: a cluster analysis that identifies four categories of extreme poverty and a w ithin-cluster ranking of municipalities (FLACSO, 2005). The cluster analysis attempts to combine structural and current dimensions of poverty. In order to do so, it uses tw o variables: 1) the proportion of 1st grade students w ith severe height-for-age deficit (stunting); and 2) income poverty incidence (the poverty headcount). W hile the first indicator points to persistent deficits in consumption that result in the accumulated effect of malnutrition of children, the second one reflects a more immediate and fluctuating aspect of deprivation. Based on these indicators, municipalities w ere grouped into the follow ing categories: • Very high extreme poverty, comprising 32 municipalities; • High extreme poverty, comprising 68 municipalities; • Moderate extreme poverty, comprising 82 municipalities; • Low extreme poverty, comprising 80 municipalities.

The 100 municipalities that are characterized by very high extreme poverty and high extreme poverty w ere identified as the target of Red Solidaria. How ever, the phased implementation of the programme required some kind of ranking among municipalities included w ithin the same cluster. For this purpose, the poverty map estimates an Integrated Index of Municipal Marginality (IIMM), w hich has a multidimensional approach to poverty. International Poverty Centre 11

The IIMM combines and w eights the monetary aspects of deprivation, measured by the income poverty gap of each municipality, w ith indicators related to unsatisfied basic needs in housing and education. These latter include: lack of access to piped w ater; lack of access to sanitation and electricity; crow ding; prevalence of earthen floors; illiteracy and low schooling rates. The sequenced inclusion of each targeted municipality in Red Solidaria has carefully respected the poverty map ranking, rather than being guided by supply constraints or political considerations. The first 32 municipalities classified under the category of very high extreme poverty w ere the first to be included, according to their IIMM position. In 2007, the programme is expanding its reach tow ards the second cluster of 68 municipalities characterized by high extreme poverty, also follow ing the IIMM ranking.

6.1 CATEGORICAL TARGETING For household targeting, different criteria have been applied. In the rural areas of the 32 municipalities that belong to the first cluster (very high extreme poverty), every family w ith pregnant w omen and children under 15 years old w ho have not finished 6th grade has been selected as a potential beneficiary. In urban areas, besides this demographic criterion, the follow ing parameters have been included: housing conditions (material used on roof, w alls and floor), crow ding and access to sanitation.13 The rationale underpinning categorical targeting is the pervasiveness of poverty in these municipalities, particularly in rural areas.14 More focused household targeting in this context w ould mean unjustifiable administrative costs and, hypothetically, could lead to social tensions w ithin small, isolated communities, as reported in the first evaluations of the Mexican pioneer CCT (Adato et al., 2000). Categorical targeting has the additional advantage of using very transparent and straightforw ard eligibility criteria, a feature that might be lost w ith the use of multidimensional indices of w ell-being or complex and secretive formulas for estimating household income through proxy means tests (PMT). Based on recent evaluation reports on the Nicaraguan and Mexican CCTs, Handa and Davis (2006) stress, how ever, the difficulties of maintaining a categorical approach as the programme expands tow ards better-off communities and requires more accurate mechanisms to identify the poorest. In this sense, for the 68 municipalities belonging to the second cluster, the Red’s operational manual details a proxy means text that w ould combine variables related to the household’s location (rural or urban), housing conditions, possession of durable goods, family composition and characteristics, and access to remittances from abroad. During fieldw ork, how ever, there appeared to be some uncertainty regarding the use of the PMT, at least based on its original formula. Further studies to test its validity and applicability w ere to be carried out on FISDL’s databases before a final decision w as to be taken on how to tackle household targeting in those municipalities.15

6.2 HOUSEHOLD CENSUS The mechanism that underpins the process of household targeting is a household census carried out by the Social Investment Fund for Local Development (FISDL) in every target municipality, after a detailed cartographic study is used to identify and locate all households in that municipality. 12 Country Study nº 9

Cartographic studies are contracted out to NGOs or private institutions, w hose w ork is closely supervised by FISDL. There is a strong component of local participation in this process, as representatives of the municipality and community leaders are closely involved in the initial identification of the population of the different rural districts and communities of that area (cantones and caseríos), as w ell as in the validation of the draft maps produced. Also, except for the fieldw ork supervisor, all cartographers must be recruited at the municipality level and receive specific training for their w ork. The cartographic study maps all geographical landmarks and physical infrastructure found in the municipality w ith GPS equipment. Its final products are detailed maps that encompass urban and rural areas and provide information at the level of small rural communities, as w ell as a directory of households by community. Besides being used by FISDL as an input for the census of potential beneficiaries, the maps are handed back to the local level through the Red’s municipal committee. As such, they may be used for the planning and implementation of other public policies and investments in the municipality.16 Using the cartographic study, a thorough household census is carried out in each municipality to be included in the Red. The census covers both rural and urban areas and uses a detailed questionnaire on housing conditions, socio-demographic characteristics of all household members, w orking activities of household members older than 10 years, as w ell as other general household information. It also includes a declaration of the respondent that all the information given is true and that the government is authorized to use it in directing social policies. It seems that a substantial portion of the information captured by the questionnaire is not used by the system that screens potential beneficiaries on the basis of the census database. Some of the questions are related to the formula for proxy means testing, but others are not even related to that aspect. Thus, the reasons for their inclusion in the questionnaire are unclear. Some of them could be important inputs for other government programmes, if the census database w ere shared w ith different organizations—a development that has not happened so far. For instance, in the case of children w ho are not attending school, the questionnaire collects information on the reasons. This could be a strategic guide for the interventions of the Ministry of Education. Another example is information related to indigenous ancestry. Most of the programme managers credit the inclusion of this aspect to the international donors involved in the design of the programme, and not to a specific issue or concern of the programme itself or even the government of El Salvador.17 The census process w as originally contracted out. Due to problems of data quality, it has now been completely undertaken by the Social Investment Fund for Local Development (FISDL), w hich directly hires, trains and supervises the researchers and the data-entry personnel. This requires considerable field-w ork, for w hich FISDL has made significant investments, including the purchase of equipment and vehicles. Once the census is completed and the data are entered in the database, a list of potential beneficiaries in every target municipality is created. It is in this stage that the verification of the additional parameters applied to urban families takes place, as urban potential beneficiaries International Poverty Centre 13

are revisited by members of the Red’s municipal committee (usually FISDL’s local representative for that municipality or the municipal liaison to the programme). According to the programme managers, because of Red’s rural emphasis, the proportion of urban beneficiaries has been kept low . Nevertheless, it seems that there w ould be no need for revisiting these families since the additional parameters could all be derived from census data. It is possible that some other discretionary parameters have been applied during these visits, but further evidence w ould be needed to confirm or deny this hypothesis.18

6.3 INCLUSION AND EXCLUSION ERRORS The actual list of eligible households by municipality is generated at the central level and the NGOs for family support and follow -up promote incorporation events for the signing of the family agreement. There are no estimates of targeting errors available. According to the statistics of the poverty map, around 75 per cent of the households of the 32 municipalities of the first cluster are classified as poor or extremely poor, w hile the Red’s CCT component covers around 70 per cent of the households in these same municipalities. Thus, under-coverage is presumably low . Since the programme adopted categorical targeting in the rural areas of these 32 municipalities, possible inclusion or exclusion errors w ould be identified mostly in urban areas due to the implementation of the special parameters applied, or w ould be related to failures in the cartographic studies or in the census process. How ever, it should be noted that the static nature of these processes (w hich are the only w ay to enter the programme) might end up also generating exclusion errors, due to the demographic dynamics of the households or family migration to targeted municipalities. Regarding inclusion errors, programme managers have pointed to queuing to collect the transfers as an additional self-targeting mechanism that is able to minimize leakage to the non- poor. But the opposite argument could be made: transaction and opportunity costs incurred to collect the transfers could favour the inclusion of the better-off, w ho can afford such costs. Targeting errors might be aggravated w ith the use of the PMT to identify eligible beneficiaries in the municipalities from the second cluster. Further analyses of the census database w ould be required to examine this issue in detail.19 Nevertheless, international experience indicates that targeting accuracy is much more an issue of implementing capacity, institutional arrangements and monitoring and oversight mechanisms than an inherent characteristic of specific targeting methods (Coady et al 2004; Castañeda et al 2005).

7 INSTITU TIONAL STRU CTU RE AND IM PLEM ENTING CAPACITY

At the national level, Red Solidaria comprises the follow ing institutional structure. The political coordination of the programme falls under the Presidency, through the coordination of the social cabinet by the Technical Secretariat. Additionally, there is the Directive Council of the Red, w hich gathers around 20 different government organizations related to the programme at ministerial level (including the Minister of Finance). This Council is supposed to meet tw ice a month and has the role of ensuring the strategic coordination required for the implementation of the Red as w ell as of other policies directed tow ards extreme poverty in El Salvador. 14 Country Study nº 9

The technical coordination of the programme is charged to an Executive Directorate, w hich w orks in close coordination w ith the Social Investment Fund for Local Development (FISDL), the Red’s implementing agency. The Executive Directorate has a very slim structure (currently, four employees plus the Director) and is hierarchically linked to the Presidency, although physically located at FISDL headquarters. The inter-institutional coordination required at the technical level takes place at an Intersectoral Committee, w here the organizations represented in the Directive Council have a seat. The rationale for these inter-institutional structures is related not only to the integrated approach of a CCT programme, w hich combines cash transfers w ith co-responsibilities in health and education, but also to the other components of the Red that deal w ith infrastructure investments, productive projects and micro-credit schemes. FISDL coordinates and supervises the programme’s operations. It is the successor of the original Social Investment Fund (FIS), created in the early 1990s w ith the same characteristics as the social fund model that prevailed in the Latin American development agenda of that time: a temporary mandate, a focus on infrastructure projects, a demand-based orientation and a heavy reliance on external resources. During the 1990s, the Fund achieved permanent status and underw ent several transformations. Since 1997, its mission has been the “eradication of poverty in El Salvador” and its mandate has included a strong component of local development and coordination w ith municipal governments. FISDL’s history, mandate and nation-w ide netw ork, through its 44 local representatives, made it a logical choice as the implementing agency of Red Solidaria, w hich is now its biggest and most important programme.

7.1 LOCAL COORDINATORS The implementation of Red Solidaria is coordinated by FISDL’s local representatives in charge of the targeted municipalities,20 w ho w ork closely w ith tw o other key actors: the municipal liaison to the programme and the NGO for family support and follow -up. Municipalities get formally involved in the Red through signing an agreement betw een the mayor and FISDL, w hich includes the commitment to support the Red’s local activities and the appointment of a municipal liaison for the programme. This latter representative is usually the mayor himself or another employee or elected representative appointed by the mayor for this task. As is the case w ith the FISDL local representative, the municipal liaison’s involvement w ith the Red covers its three components and is an add-on function, for w hich there is no specific remuneration or incentive. The NGOs for family support and follow -up are hired annually by FISDL through a bidding process for the implementation of the CCT component in each municipality. During the research fieldw ork, there w ere five NGOs w orking w ith the first 32 municipalities included in the programme and the bidding by NGOs that w ould w ork w ith the 15 municipalities to enter the programme in 2007 w as underw ay. The NGOs have a regional coordinator, a municipal coordinator and a team of local promoters, usually young leaders recruited in the municipality, w ho are each responsible for w orking w ith 150-160 families. In practice, they are the main link betw een beneficiaries and the programme. International Poverty Centre 15

As described above, the NGOs are in charge of the bulk of the programme’s local activities, including: organizing events for the signing of family agreements; monitoring and reporting of the compliance w ith co-responsibilities; conducting lifelong learning sessions for beneficiary families; coordinating (along w ith the FISDL local representative) payment events; receiving complaints; and administering and updating the information that feeds into the beneficiary registry system. Most of this w ork requires frequent visits to beneficiary households, but local promoters interview ed during the research pointed out the difficulties in reaching the remotest communities—especially since most of them have to reach these destinations on foot. These three key stakeholders (FISDL local representatives, municipal liaisons and NGOs for family support and follow -up) are the core of the municipal committees of the programme, w hich also include representatives from the health and education sectors (usually the director of the health unit in the municipality and an employee of the departmental office of the Ministry of Education), as w ell as community leaders and, in some cases, representatives of other government agencies w orking w ith the Red in that particular tow n (e.g., Ministry of Agriculture and National Registry Agency). The FISDL local representative is responsible for the coordination of the committee and its regular meetings, w hich should take place monthly. The municipal committee’s main role is the inter-institutional coordination at the local level for the implementation of the Red’s three components. A substantial portion of its mandate is related to the infrastructure component through the selection and oversight of construction and rehabilitation projects. In the CCT component the committee’s activities include: validation of the cartographic study; follow -up and validation of the census of potential beneficiaries; validation of urban families that comply w ith the special targeting parameters; and monitoring of updates and exclusions w ith regard to the list of beneficiaries that are proposed by the NGO for family support and follow -up. Social monitoring and transparency are thus one of the objectives of the municipal committees since they can assess the exclusions of beneficiaries proposed by the NGOs, function as a channel for complaints from beneficiaries and non-beneficiaries and provide an alternative channel to the NGOs.

7.2 STRENGTHENING LOCAL PARTICIPATION Intended or unintended attempts to forge new accountability relationships betw een central and local governments, as w ell as betw een service providers and beneficiaries are frequently present in CCT programmes (de la Brière and Raw lings, 2006). Local instances of participation are usually included in the programmes’ strategy to foster community participation or civil society engagement that contributes to greater transparency and efficiency. The reason is that they can combine functions of direct accountability (as social control mechanisms) and voice (as channels for community-based or beneficiary demands) (de Janvry et al, 2005). Nonetheless, the extent to w hich they are effectively put in place varies according to the degree of decentralization experienced in implementation, as w ell as being dependent on institutional settings and contextual factors related to local governance and politics. Furthermore, depending on how they are formed and function, their performance can be limited to merely legitimizing decisions taken elsew here. 16 Country Study nº 9

In Red Solidaria, there are shortcomings in the functioning of municipal committees. Although they have been officially set up in the 47 municipalities that the Red is expected to cover by the end of 2007, their w ork has not been homogeneous. It depends heavily on the engagement of its members, particularly the FISDL local representative, w ho makes sure that the committee meets regularly and coordinates its meetings. Some of the committees’ roles, such as the validation of urban families as beneficiaries, have been taken over by a few of their members, such as the FISDL representative and the municipal liaison. One bottleneck seems to be the lack of participation of community leaders, w ho receive no direct incentives to attend the meetings, having to cover long distances, usually on foot, and lose a day’s w ork to be present. And it is not clear the extent to w hich the community leaders included in the committees indeed represent beneficiaries. A common feature of these leaders is that most of them are males, an outcome that might indicate the prevalence of a cultural gender bias in their choice. Besides the municipal committee, the Red’s local institutional design involves a beneficiary or community committee, w hich should comprise at least three beneficiaries in each district (canton), w ho function as liaisons betw een the community and the NGO for family support and follow -up. It seems, how ever, that they have not been set up in most communities. Throughout the field research, it w as not possible to find evidence related to their w ork. It should be noted that programme managers are aw are of these shortcomings and have developed a detailed operational guide for the functioning of municipal committees, as w ell as a monitoring system that includes in-person oversight of committee meetings and a digital platform for a virtual follow -up of their activities. The dissemination of this guide and the implementation of the monitoring system have begun in 2007. In any case, the vertical organizational structure of the programme, w hich relies on a top- dow n approach from the Social Investment Fund for Local Development (FISDL), might create additional challenges for the programme’s expansion, since it bears the potential of creating conflict, particularly in municipalities that are not governed by ARENA.

8 INTERSECTORAL COORD INATION

Red Solidaria w as launched as part of the government’s social strategy called O portunidades. This strategy is described as an effort to develop people’s capabilities and opportunities for social and economic inclusion in El Salvador. Besides Red Solidaria, targeted at the poorest rural population, this strategy comprises four other strategic areas: health (Fosalud), youth (Jóvenes), information technology (Conéctate) and credit (Tu Credito). The w ay that these five programmes interconnect is, how ever, not self-evident. Apparently, there is no direct coordination betw een them, w ith the exception of the micro- credit projects that form part of the Red’s component of family sustainability. The five components seem to have been grouped together more because of the presidential priority that they have received than being based on an articulated social strategy, as claimed. Nevertheless, one can argue that Red Solidaria does achieve an important intersectoral coordination w ith education and health, particularly at the national level. The Ministries of Health and Education have taken part in the programme since its design, including the formulation of the logical framew ork, the definition of co-responsibilities and the follow -up mechanisms. International Poverty Centre 17

In the implementation phase, these ministers are important partners in the supply-side efforts that have to be linked w ith a CCT-like programme, through the programmes ‘Extension of Health Services’ (ESS) and ‘Effective Schools Netw orks’ (REE), w hich are targeted at the same municipalities reached by the Red.21 The ministries are also involved in the impact evaluation planned to start in 2007. This coordination takes place not only at the political level, through the Red’s Directive Council, but also at the technical level, through the Intersectoral Committee. Moreover, the Red goes beyond simply establishing add-on responsibilities for the health and education personnel associated w ith its CCT component. It directs additional resources for strengthening education and health infrastructure through its second component. Thus, the Red has not led only to an additional burden or a reorientation of line ministries’ funds for these areas. It has also meant an infusion of ‘new money’ and potentially greater impacts for the ministries’ ow n programmes. At the municipal level, the coordination of health and education providers w ith Red Solidaria is related mostly to feeding the information flow needed for the verification of co-responsibilities that are monitored by the NGOs for family support and follow -up. Interview s w ith the service providers visited during the research fieldw ork indicated that the degree of familiarity w ith the programme appeared higher among health personnel than education officers. This could be related to the fact that the director of the municipal health unit has a seat on the municipal committee, w hile the education representative comes from the departmental level. Even w ith a fair degree of intersectoral coordination and an integrated design, there have been supply-side constraints on the programme. For instance, new spaper articles have pointed out that the beginning of the school year after the Red w as introduced in the first 15 targeted municipalities w as marked by shortages of classrooms, equipment and teachers needed to respond to the increases in enrolment resulting from the programme. Officials of the Ministry of Education confirmed these difficulties, but pointed out they are commonly observed in many urban schools of the country. The Red’s novelty w as that it focused on poor and remote rural areas, often ‘invisible’ to regular social programmes—and that a greater speed in government response w as required so as to avoid jeopardizing the programme’s co-responsibilities, w hich have been a highly valued aspect of the Red. In this sense, incentives that increase the demand for social services, especially w hen they are linked to such a visible programme as Red Solidaria, might w ork as positive pressures for the expansion of service coverage and motivate the population to access basic rights to w hich they are entitled.

8.1 DIRECT AND INDIRECT COSTS An important issue, w hich does not appear as a specific education sector programme but is clearly linked to the Red’s success, is school transportation. As most of the Red’s beneficiaries are located in rural areas, often unattended by public transportation and lacking roads in adequate condition, school distance can be an obstacle for children’s attendance. Apparently, each school netw ork incorporated into the REE programme could use part of its money to set up transportation schemes. How ever, there is no national programme that addresses this potential supply-side constraint and REE funds might not be sufficient to provide an appropriate solution. 18 Country Study nº 9

In the health sector, increases in demand resulting from Red Solidaria have also required hiring more doctors and nurses and finding other ‘creative’ approaches, such as directing to the Red-covered municipalities medical students that are fulfilling their one-year social service requirements. W hat remains to be further investigated is w hether these responses address only the immediate objective of ensuring that beneficiaries comply w ith the programme’s co-responsibilities but do not offer them quality services in education and health that go beyond meeting the formal requirements. Q uality of service provision is an issue of concern pointed out by many analysts w ho have examined supply-side constraints that might jeopardize the long-term effectiveness of CCTs (Britto, 2005; Handa and Davis, 2006; Cohen and Franco, 2006; de la Brière and Raw lings, 2006). In the case of education and health in El Salvador, the W orld Bank-led poverty assessment provides some interesting findings (W orld Bank, 2005). For instance, reasons mentioned by poor families for having children out of school related to access and affordability issues associated w ith direct and indirect costs of schooling, in addition to a lack of interest in school in some cases. This might reflect low perceived returns to education associated w ith problems of educational quality. But a general lack of availability of schools w as not w idely reported (although this might be a problem in some circumstances, such as secondary schooling in rural areas). In health, there is an affordability problem related to the practice of charging user fees in health centres—although they had been legally abolished in 2002. Furthermore, despite the main focus of the Red’s health co-responsibility on primary health care, w hich is an important issue for poor families, there is an urgent need to go beyond preventive measures to broaden access to general health insurance. More than 80 per cent of the Salvadoran population is uninsured—a proportion that reaches over 97 per cent among the extremely poor. Besides health and education, the Red’s sustainability component encompasses another relevant requirement of intersectoral coordination. This component’s productive projects are implemented by the Ministry of Agriculture (MAG), and its micro-credit schemes are under the responsibility of the Multi-Sectoral Investments Bank (BMI). Although these initiatives might coincide, in terms of geographical targeting, w ith the CCT component, their beneficiaries are not necessarily the same households that receive the cash transfers due to differing eligibility requirements and targeting criteria. Furthermore, the productive component of the Red lacks adequate resources: it has relied mostly on budget re-orientation tow ards the targeted municipalities rather than the provision of ‘new money’. The reason might be related to the w eak emphasis on promoting ‘exit doors’ for the CCT component that are not demographically based. Another possibility, mentioned during an interview , is that transferring cash or building infrastructure is more visible and less complex than promoting time-intensive productive projects that can be capable of positive and sustainable impacts over the longer term.

9 PROGRAM M E COSTS, POLITICAL SU PPORT AND SU STAINABILITY

W e could not obtain detailed information on the disaggregated budget and costs of Red Solidaria. W hen the programme w as created, estimates projected costs of around US$ 50 million per year, covering its three components. This amount is mentioned not only in the International Poverty Centre 19

conceptual document, but also in the legal framew ork of the programme. Managers have pointed out that this estimate has been maintained on average, although there have been annual variations related to phased implementation, fine-tuning and adjustments, as w ell as upfront fixed costs incurred in the design and setting-up of operations, such as the purchase of equipment and vehicles for fieldw ork. According to programme managers, the CCT component comprises around 20 per cent of the Red’s total budget, but it is not clear how much is spent on transfers and how much is absorbed by the administrative costs involved in targeting mechanisms, delivery of transfers and programme administration. These costs comprise FISDL expenditures and the costs incurred w ith the NGOs for family support and follow -up. Up until now , the sustainability component has absorbed only a small portion of the Red’s budget: less than US$ 1 million annually. As a result, the bulk of the programme’s resources are addressed to the infrastructure component. A remarkable feature of the programme is that transfers are completely financed through national funds. As much as Red Solidaria depends heavily on external resources, these have been directed to the infrastructure component and, to some extent, to the programme’s management. This might increase the prospects for the sustainability of the CCT component if one considers, in particular, that the political polarization of the country hindered the approval of foreign loans in the National Assembly.

9.1 EXTERNAL FINANCING The Inter-American Development Bank (IDB) and the W orld Bank have been important partners in Red’s conception and initial design and have negotiated loans amounting to US$ 57 million and US$ 21 million respectively, mostly to finance basic infrastructure and improvements of health supply in the programme’s targeted municipalities, as part of Red’s Component 2. Nevertheless, these loans faced strong objections from the opposition party and the deadlines for the beginning of their disbursement have not been met. Since these loans have not been approved in a timely manner, the government of El Salvador does not seem to be counting anymore on them. The programme w as able to fill in this resource gap through international cooperation grants from the European Union (totalling 37 million euros), Luxembourg (20 million euros) and Spain (10 million euros). The German bilateral cooperation agency has also had an important role in financing capacity-building activities for the Red’s implementing agency. Thus, the non-approval of foreign loans for Red Solidaria does not seem to have jeopardized, so far, the programme’s continuity. In many interview s carried out at the grassroots level w ith beneficiaries, municipal personnel, and health and education personnel, w idespread concerns about the continuity of the programme due to the non-approval of the loans w ere evident. In general, local stakeholders w ere not aw are that the CCT component did not depend on external money and that foreign grants could fill in the gap created by the absence of the loans. In this sense, bottom-up pressures could have been used to increase the prospects of approval of the W orld Bank and IDB loans. How ever, the political polarization of the country and the long history of foreign dependence and external debt appear to have been prominent factors influencing congressmen’s deliberations. 20 Country Study nº 9

9.2 POLITICAL CHALLENGES Red Solidaria originates from a presidential initiative and relies heavily on this back-up. This gives the programme considerable support and commitment from top government officials and smoothes out implementing difficulties deriving from intra-governmental conflicts. Furthermore, it makes the programme highly visible to beneficiaries and to society as a w hole. This strategy has been deliberately adopted by the government through considerable investments in publicity and the frequent association of the programme w ith the presidential image. This high programme profile also entails difficulties, how ever, in the sense that the close links w ith the president and the party in pow er make the programme a preferred target for the opposition. The difficulties in the approval of the loans for Red Solidaria illustrate this point. As could be expected, debates in the National Assembly about the loans have been politically motivated and have not centred on conceptual or design aspects of the programme itself, but rather on the allegedly propagandistic use of it, as w ell as the deepening of an already unsustainable foreign debt. Despite the political difficulties in the National Assembly, the solid design of Red Solidaria seems to have given it a considerable amount of technical legitimacy. Geographical targeting through the poverty map prevents discretionary politically-driven expansions. This is enhanced by the fact that the poverty map w as developed by a reputed academic institution (FLACSO), w ith the objective of providing inputs to FISDL’s activities in general, not just the Red. Only after the map w as ready w as the Red created and based on it. Furthermore, in 2006 new municipal governments w ere elected and took office for a three-year term. This changed the political map of the country and required new terms of agreement betw een the national government and targeted municipalities for the Red’s implementation. Despite these problems, the poverty map continued to be respected in the programme’s expansion strategy. Now , of the 47 municipalities that w ill be part of the Red until the end of 2007, 26 are ruled by the president’s party (ARENA) and 12 by its main opposition party (FLMN), w hile the nine others are governed by smaller parties. The experience of other Latin American countries w ith CCTs served as an inspiration to Red Solidaria, a fact that has been used to strengthen its technical legitimacy. The conceptual document, the programme’s w ebsite and even the poverty map reports refer to successful CCTs implemented in Chile, Colombia and Mexico. They provide evidence of positive impacts of this kind of programme and justify its introduction in El Salvador. Moreover, consultants from some of these countries have been directly involved in Red’s initial design. The Red has also been able to attract strong donor support. In spite of the controversy regarding the approval of W orld Bank and IDB loans, the programme managed to raise substantial grant funding from bilateral donors and the European Union. These countries are highlighted as strategic partners of the Red in publicity material and on the programme’s w ebsite. This support is used as a kind of guarantee of Red’s sound technical quality. Another issue that seems to contribute to the programme’s political support and sustainability is the fact that, since the design stage, line ministries have been politically and technically involved w ith it. Not only is this important in ensuring that a demand-based incentive, such as a CCT, is coordinated w ith supply-side programmes; but it also helps avoid or overcome bureaucratic jealousy. International Poverty Centre 21

In general, our impressions of the programme derived from interview s and informal talks w ith local stakeholders as w ell as new spapers articles w ere quite positive. Even though the limited scope of the research and the fieldw ork did not allow a systematic presentation of evidence in this respect, the main criticisms of the programme have been confined to specific aspects, such as the size of the transfers (judged by some to be too small and insufficient to make real impacts on the lives of people), supply-side issues (w hich have not been adequately tackled) and a supposedly propagandistic use of Red Solidaria by the government.

10 CONCLU SION: PEND ING ISSU ES AND REM AINING CHALLENGES

Red Solidaria is an example of how a small country, w ith limited resources, can design and set up a complex CCT programme. It has draw n valuable lessons from previous CCT experiences elsew here; counted on a technically strong and motivated team; received political championship from the top levels of government; set up inter-institutional mechanisms for administrative coordination; and draw n on technical and financial support form varied international sources. El Salvador’s specific features, especially its small size combined w ith an effective national institution w ith a w idespread netw ork and a mandate related to poverty reduction might have favoured the implementation strategy for Red Solidario. The programme has relied heavily on on-site supervision by FISDL personnel. Still, there are important issues and challenges related to the programme’s agenda, some of w hich are w idely acknow ledged by programme managers and identified as areas of attention in the near future.

10.1 LOCAL PARTICIPATION First of all, there is a need to strengthen the mechanisms of local participation, in particular the participation of municipal committees and beneficiary committees. They are regarded as vital arenas for ensuring the legitimacy and publicity of the Red’s crucial activities, such as targeting and programme follow -up. Even though these committees have been fulfilling these roles, greater efforts are needed to genuinely involve and give voice to representatives of the beneficiaries themselves. Much of the w ork carried out by the committees is concentrated in the hands of local FISDL representatives and municipal coordinators from the NGOs for family support and follow -up. The former handle most of the strategic instruments and information related to the infrastructure component of the Red, w hile the latter handle practically all formal contacts betw een beneficiaries and the programme (including fielding complaints and investigating justifications for non-compliance w ith co-responsibilities). Capacity-building and incentives at the grassroots level might be as important as the monitoring approach adopted by the programme’s managers in broadening the scope of participation in these committees. It w ill be important to avoid both a concentration of discretionary pow ers in the hands of these representatives and the overburdening of beneficiaries and community leaders w ith additional tasks that have only modest practical impacts. 22 Country Study nº 9

A minor issue related to the institutional capacity of the programme relates to the prevailing human resource profile in FISDL. Since it has traditionally been an organization oriented tow ards infrastructure (and the size of the infrastructure component of the Red is quite large), most of FISDL’s personnel has a background in architecture, engineering or other such sciences. Certainly this need not be an obstacle that hinders their active and enthusiastic involvement in a CCT programme, but such a profile poses additional requirements in terms of training and capacity-building, especially for the local representatives w ho w ork at the grassroots level. An important challenge for the integrated approach of the programme concerns strengthening the links betw een the CCT component and the sustainability component. Even if the Red does not emphasize productive projects as the ‘exit door’ of the transfer programme—w hich seems to be an appropriate view —Component 3, w hich covers productive projects and micro-credit schemes, still receives very small funding compared to the other tw o components. W hat is more serious is that there are no concerted efforts to ensure that this component’s beneficiaries coincide w ith the CCT beneficiaries, presumably the poorest ones in the communities.

10.2 PROGRAMME DURATION Some design issues regarding the CCT component also pose important problems. For instance, as the programme has an expected duration of three years for each beneficiary family, it does not cover the time needed for the human capital impacts intended by the Red. The w hole monitoring system for children’s health and immunizations covers children up to the age of five years. This means that a beneficiary family w ith a pregnant w oman w ould have to be in the programme for at least six years to have the entire maternal and child health protocol covered. In the case of education, limiting the programme’s grants to families w ith children w ho have not completed 6th grade not only contradicts the goal of ensuring at least basic education to all, but might also create negative incentives for grade promotion among students w ho do not have the relevant age for their grade. W ithout any problems of repetition or drop-out, students w ould reach 6th grade at the age of 12, not 15 years. Strengthening the delivery of basic education beyond 6th grade in the targeted municipalities is needed to remove this potential perverse incentive from the programme, since supply constraints have been pointed out as the main reason for the grade-related eligibility requirement. Dealing w ith the quality of education and providing more than just preventive health care for the poor are also important areas that need further attention.

10.3 TARGETING CRITERIA W ith regard to accountability issues, there is a lack of transparency in the targeting criteria used by the PMT, w hose actual formula still seems to be under discussion. On the one hand, the application of an econometric formula to the census data to determine eligibility is likely to make it more difficult for potential beneficiaries to enhance their chances of being included in the programme (if they are unaw are of the scoring w eights being used). On the other hand, it might generate perceptions of unfairness that cannot be easily addressed because the eligibility requirements are not easily understood. International Poverty Centre 23

To some extent, these perceptions have already developed as a result of the application of the additional targeting parameters for urban areas. They might also w orsen as the programme moves tow ards covering the broader population identified as the second cluster through the use of the poverty map. The same problem affects the co-responsibilities that might not be, in reality, conditionalities for receiving the cash transfers. Participation in family training sessions is frequently presented to beneficiaries as a conditionality, similar to school attendance and health check-ups. How ever, even though such sessions might be an important element of the CCT component, for the sake of transparency they should be presented as w hat they really are: co-responsibilities that do not entail discounts in the monthly grants.

10.4 ENTRY DOORS Another area of concern relates to the static approach of the programme in the w ay that it identifies and includes beneficiaries. Since targeting is completely based on a census that takes place only once, it does not take into account poverty or demographic dynamics. The only continuous updating of the list of beneficiaries concerns cases of exclusion, w hich arise due either to the correction of previous inclusion errors or situations in w hich beneficiaries exceed the eligibility thresholds (e.g., children turning 16 years of age or finishing 6th grade). More dynamically based ‘entry doors’ to the programme, w hich are also carefully designed not to create fertility incentives, could be developed to address this limitation. But this w ould make sense only if the programme overcomes its temporary status, i.e., it is extended beyond the year 2009. In this sense, Red Solidaria w ould have to be institutionalized as a permanent programme and be complemented by other programmes that are part of a coherent social protection strategy. As much as the programme might achieve important impacts on the w ell being of poor rural households w ith children, many other vulnerable groups in El Salvador remain uncovered by any kind of social assistance or social insurance scheme. Thus, an integrated social protection strategy has to consider how to reach the urban poor, the elderly and the disabled, taking account of El Salvador’s specific context of pervasive informality in the labour market, serious fiscal constraints and political polarization. This is a challenge that w ould require designing various kinds of social programmes—a task that w ould be much more complex than simply expanding the CCT model itself. 24 Country Study nº 9

APPEND IX A

RESEARCH FIELDW ORK Mission to El Salvador: 19-28 March, 2007 Interview s

Organization, representative Local Health unit, Director Primary school, Director Town hall, Red Solidaria municipal liaison National FISDL, Research and development manager FISDL, Research coordinator FISDL, Red Solidaria registry and transfers coordinator FISDL, Red Solidaria local coordinator FISDL, Census coordinator FISDL, Cartography coordinator Ministry of Agriculture, Red Solidaria’s focal point Ministry of Education, Coordinator of REE Ministry of Education, Red Solidaria’s focal point Ministry of Health, Red Solidaria’s focal point Red Solidaria, Executive director Red Solidaria, Local development expert Red Solidaria, Intersectoral manager Red Solidaria, Social visibility coordinator Technical Secretariat of the Presidency, Coordinator of the social cabinet International IDB, Social programmes division expert (Washington, DC)

Field visits

Municipality, Poverty map Poverty map Activity Department ranking cluster Observation of FISDL census, informal talks with Jicalapa, High extreme 34 household members, census personnel (supervisor La Libertad poverty (2) and researchers) and FISDL local representative Very high Visit to health unit and primary school, informal talks Jutiapa, Cabañas 8 extreme poverty with beneficiaries, interviews with local stakeholders (1) Very high Observation of a municipal committee session Masahuat, Santa 20 extreme poverty Ana (1) Observation of Red Solidaria payment event, Very high San Agustín, informal talks with beneficiaries, NGO local 24 extreme poverty Usulatán promoters, health unit personnel and FISDL local (1) representative International Poverty Centre 25

APPEND IX B

Targeted municipalities

% Extremely Number of % Poor Department Municipality Population Cluster poor households households households

1 Morazan 342 1,801 1 88.6 60.5 San Antonio del 2 San Miguel 1,126 6,373 1 78.1 59.5 Norte 3 Sonsonate Cuisnahuat 2,356 12,854 1 76.1 52.4 4 Morazan 2,343 10,907 1 78.2 53.1 5 Sonsonate Caluco 1,484 8,341 1 79.1 43.7 6 Usulatan Nueva Granada 1,604 8,307 1 84.4 58.9 7 Chalatenango San Fernando 529 2,819 1 81.9 59.7 8 Cabanas Jutiapa 1,256 7,178 1 77.8 53.9 9 Morazan 580 3,090 1 80.3 56.6 10 San Miguel Carolina 2,011 10,215 1 67.9 45.6 11 Morazan San Isidro 649 3,543 1 73.7 49.8 12 Cabanas Cinquera 255 1,184 1 83.1 52.6 San Jose 13 Chalatenango 421 1,915 1 81.2 58.0 Cancasque 14 Morazan 838 4,176 1 74.0 43.2 15 Ahuachapan Guaymango 3,921 19,832 1 72.7 47.2 San Isidro 16 Chalatenango 83 376 1 75.9 59.0 Labrador San Francisco 17 Chalatenango 518 2,718 1 77.2 47.9 Morazan 18 Morazan San Simon 1,567 8,062 1 70.1 44.2 19 Santa Ana Masahut 801 4,113 1 77.8 48.2 20 Chalatenango Arcato 590 2,895 1 72.0 48.3 21 San Vicente Santa Clara 897 4,632 1 73.8 47.8 San Antonio 22 Chalatenango 334 1,485 1 80.2 49.7 Ranchos San Esteban 23 San Vicente 906 4,598 1 75.7 43.1 Catarina 24 Usulatan San Agustin 932 4,448 1 66.3 46.6 Santiago de la 25 Santa Ana 1,539 6,589 1 68.2 44.6 Frontera Santo Domingo 26 Sonsonate 1,471 7,278 1 71.7 44.5 de Guzman 27 Chalatenango La Laguna 993 5,111 1 71.1 47.1 Paraiso de 28 La Paz 639 2,933 1 76.4 44.8 Osorio 29 Chalatenango Ojos de Agua 773 3,536 1 71.8 50.6 30 Chalatenango Las Vueltas 418 2,045 1 75.1 45.7 31 Usulatan Estanzuelas 2,097 10,102 1 68.2 44.3 32 Chalatenango Potonico 372 1,582 1 76.3 47.3 26 Country Study nº 9

Nuevo Eden de 33 San Miguel 623 3,160 2 62.8 39.5 San Juan 34 La Libertad Jucalapa 1,142 6,056 2 63.9 41.6 35 La Union Lislique 2,700 13,790 2 63.6 37.0 36 Cabanas Dolores 1,182 6,143 2 64.6 37.9 Santa Isabel 37 Sonsonate 1,885 9,790 2 60.6 40.8 Ishuatan 38 Morazan 2,044 9,351 2 68.5 40.9 San Francisco 39 Usulatan 1,439 6,058 2 72.6 40.6 Javier 40 Morazan Corinto 3,129 16,402 2 65.1 35.6 41 Morazan 807 3,880 2 63.6 38.8 42 Ahuachapan Jujutla 5,655 28,059 2 59.9 35.7 43 San Miguel Sesori 2,263 11,379 2 54.8 32.1 44 Ahuachapan Tacuba 4,529 22,425 2 68.1 42.2 Monte San 45 Cuscatlan 1,665 8,973 2 64.3 34.8 Juan 46 Morazan 427 2,129 2 58.3 31.2 47 Morazan Chilanga 1,715 8,358 2 65.5 38.7 48 Morazan San Fernando 373 1,924 2 66.8 38.1 49 Morazan 692 3,127 2 63.7 40.3 Santa Catarina 50 Sonsonate 1,702 8,635 2 74.5 34.6 Masahuat 51 La Libertad Teotepeque 2,730 14,108 2 58.5 34.3 52 San Miguel San Gerardo 1,093 5,579 2 54.8 33.6 53 Chalatenango Nueva Trinidad 342 1,896 2 65.5 40.4 54 Morazan 1,020 4,867 2 63.2 33.7 55 Cuscatlan San Cristobal 1,337 7,130 2 60.5 31.3 56 San Vicente San Ildefonso 1,811 8,704 2 60.4 36.1 San Pedro 57 Ahuachapan 1,662 8,411 2 70.6 39.4 Puxtla 58 La Libertad Comasagua 2,418 11,155 2 60.3 35.8 59 Chalatenango Agua Caliente 1,753 8,027 2 57.3 32.9 60 La Libertad Chiltiupan 2,172 10,988 2 55.2 36.3 61 Cabanas Victoria 2,790 14,796 2 65.9 42.1 62 Usulatan Alegria 2,540 13,813 2 74.1 36.8 63 Usulatan Tecapan 2,593 11,475 2 64.5 35.5 San Antonio de 64 Chalatenango 386 2,113 2 64.8 42.8 La Cruz 65 San Miguel Ciudad Barrios 6,131 29,664 2 59.6 32.9 Concepcion 66 Usulutan 2,931 13,086 2 57.0 30.8 Batres 67 Usulutan Berlin 3,712 17,952 2 64.9 35.2 68 Usulutan Ozatlan 2,709 12,140 2 58.8 31.1 69 Chalatenango Las Flores 490 2,072 2 68.0 41.8 70 La Paz San Emiglio 531 2,689 2 62.1 38.8 San Miguel 71 La Paz 1,011 4,629 2 70.0 38.0 Tepezontes Santa Maria 72 La Paz 1,249 5,660 2 63.1 34.0 Ostuma 73 La Union Yayantique 1,230 5,412 2 58.4 32.0 74 San Miguel San Jorge 2,168 9,641 2 55.4 34.4 International Poverty Centre 27

Mercedes 75 Usulatan 2,933 13,866 2 60.4 34.8 Uamana 76 Chalatenango El Carrizal 561 2,757 2 71.7 40.3 77 San Vicente San Lorenzo 1,270 6,178 2 66.1 37.9 San Juan 78 La Paz 758 3,190 2 64.4 36.0 Tepezontes 79 San Vicente Apastepeque 3,541 18,247 2 62.5 33.1 Nombre de 80 Chalatenango 871 4,341 2 62.7 36.3 Jesus 81 Usulutan Jacuaran 3,047 12,846 2 65.9 31.4 82 Cabanas Ilobasco 11,938 66,259 2 60.4 31.0 83 Morazan El Rosario 257 1,076 2 63.4 31.5 84 Chalatenango Comalapa 998 4,364 2 64.6 40.0 Santa Cruz 85 Cuscatlan 561 2,387 2 62.7 36.5 Analquito San Pedro 86 La Paz 2,017 9,430 2 59.7 32.6 Nonualco 87 Cuscatlan El Rosario 737 3,762 2 60.2 33.5 88 Cabanas Tejutepeque 1,258 5,863 2 60.5 36.0 San Jose da la 89 La Union 962 3,975 2 57.1 31.3 Fuente Delicias de 90 Morazan 1,029 4,749 2 59.9 30.1 Concepcion San Antonio 91 La Paz 831 4,047 2 60.4 33.6 Masahuat San Luis del 92 Chalatenango 501 1,700 2 65.9 37.7 Carmen 93 La Paz Tapalhuaca 907 3,612 2 53.5 32.2 94 San Vicente Verapaz 1,812 7,803 2 64.2 31.8 Santa Rosa 95 Santa Ana 1,398 6,185 2 57.6 32.6 Guachipilin 96 Usulutan Santa Elena 3,389 15,476 2 56.2 31.7 97 Cuscatlan Tenancingo 1,692 7,858 2 52.3 30.1 Concepcion 98 Chalatenango 1,482 6,387 2 61.7 31.9 Quezaltepeque San Miguel de 99 Chalatenango 569 2,438 2 56.1 34.5 Mercedes San Francisco 100 Chalatenango 262 1,067 2 57.4 31.9 Lempa Source: Flacso, 2005. 28 Country Study nº 9

REFERENCES

Adato, M. (2000). ‘The Impact of PRO GRESA on Community Social Relationships’. W ashington, DC: International Food Policy Research Institute. de la Brière, B. and Raw lings, Laura (2006). Exam ining Conditional Cash Transfers: a Role for Increased Social Inclusion? Social Safety Net Primer Series. W ashington, DC: The W orld Bank. Britto, T. (2005). ‘Recent Trends in the Development Agenda of Latin America: an Analysis of Conditional Cash Transfers’. Paper presented at the Social Protection for Chronic Poverty Conference. Manchester: IDPM, Chronic Poverty Research Centre. Castañeda, T., Lindert, K., de la Brière, B., Fernandez, C., Hubert, C., Larrañaga, O., Orozco, M., and Viquez, R. (2005). ‘Designing and Implementing Household Targeting Systems: Lessons from Latin America and the United States’. Social Protection Discussion Paper Series no. 0526. W ashington, DC: The W orld Bank. Coady, D., Grosh, M. and Hoddinott, J. (2002). ‘The Targeting of Transfers in Developing Countries: Review of Experiences and Lessons’. Social Safety Net Primer Series. W ashington, DC: The W orld Bank. Cohen, E. and Franco, R. (2006). ‘Los Program as de Transferencias con Corresponsabilidad en Am érica Latina: Sim ilitudes y D iferencias’ in Cohen, E. and Franco, R. (eds.) Transferencias con Corresponsabilidad: una M irada Latinoam ericana. Mexico: Sedesol, 2006. de Janvry, A., Finan, F, Sadoulet, E., Nelson, D., Lindert, K., de la Brière, B. and Lanjouw , P. (2005). ‘Brazil’s Bolsa Escola Program: the Role of Local Governance in Decentralized Implementation’. Social Protection Discussion Paper no. 0542. W ashington, DC: The W orld Bank. FLACSO Programa El Salvador (2005). M apa de Pobreza: Política Social y Focalización – tom o 1. El Salvador: FISDL. FUNDE (2005). GPN Global Labor M arket D atabase: El Salvador, 2005. El Salvador: Fundación Nacional para el Desarrollo. Available at w w w .gpn.org. Handa, S and Davis, B. (2006). ‘The Experience of Conditional Cash Transfers in Latin America and the Caribbean’. D evelopm ent Policy Review , 24 (5), pp. 513-536. Molyneux, M. (2006). ‘Mothers at the Service of the New Poverty Agenda: Progresa/Oportunidades, Mexico’s Conditional Transfer Programme’. Social Policy and Adm inistration, vol. 40, no. 4, August, pp. 425-449. PNUD (2003). Inform e sobre D esarrollo Hum ano – El Salvador 2003. El Salvador: PNUD. Red Solidaria (2006). M anual O perativo de la Red Solidaria de los Contratos de Prestam o BID 1665/O C-ES y BM 7338-ES. D ocum ento final: julio de 2006. El Salvador: Secretaria Técnica de la Presidencia. Red Solidaria (2007). Red en Acción – Febrero 2007. El Salvador: FISDL. Soares, F. and Ribas, R. (forthcoming). ‘Programa Tekoporã – Evaluación del marco lógico, del manual operativo y de la línea de base del piloto (Primer Informe Parcial)’. Country Study, Brasilia: International Poverty Centre. W orld Bank (2005). El Salvador Poverty Assessm ent: Strengthening Social Policy. Report no. 29594-SV. W ashington, DC: The W orld Bank. .

NOTES

1. International policy learning and external influence, especially from the W orld Bank and the Inter-American Development Bank, also played an important role in the agenda setting and technical design of the programme. 2. Executive Decree n. 11, of March 4th, 2005, modified by Executive Decree n. 42, of May 16th, 2005. 3. The Red’s legal document limits its duration to 2009 but the programme’s conceptual document includes goals for 2010-2015 and refers explicitly to the MDGs. 4. Nevertheless, the same poverty assessment show s surprising data related to recent poverty dynamics among the rural and urban populations. In the first half of the 1990s, economic grow th w as able to reduce poverty, particularly among the moderate poor and the urban poor. This process slow ed dow n after 1995 and decelerated further in the 2000s, due to domestic factors, exogenous shocks and natural disasters. This reduced considerably the pace w ith w hich the country w as able to improve the conditions of the rural poor, but the urban poor seemed to be even more affected: urban poverty actually increased betw een 2000 and 2002 (W orld Bank, 2005). This might be related to the effect of migrant’s remittances, w hich have had greater poverty impacts in rural areas (PNUD, 2003). 5. Although poverty trends in El Salvador seem to be consistent regardless of the methodology used, the statistics regarding current estimates and the speed of poverty reduction throughout the 1990s are controversial (PNUD, 2003). 6. There is some ambiguity in the use of the terms ‘poverty’ and ‘extreme poverty’ in Red’s documents. The poverty line in El Salvador is officially calculated as tw ice the costs of a basic food basket (US$ 2.12 dollars per person per day in urban areas and US$ 1.33 dollars per person per day in rural areas). The extreme poverty line is equivalent to the costs of one basic food basket (FLACSO, 2005). How ever, these values are contested and might be underestimated (PNUD, 2003). 7. W hile this is generally regarded as an empow ering tool for w omen, Molyneux (2006) w arns against overemphasizing the ‘maternalistic’ gender role in the poverty agenda epitomized by CCTs. 8. A field visit during pay day provided anecdotal evidence in this regard. Beneficiaries of a remote district complained about having to rent a truck to drive them to the payment location at daw n, w ait for all of them to receive their stipend and bring them back home. The w hole operation w ould take the w hole morning and cost around US$ 2.50 per person (an amount considered an overestimate, how ever, by programme managers). 9. Despite formally including pre-school enrolment and attendance co-responsibilities, the Red does not alw ays enforce the schooling conditionality for children younger than seven years old, due primarily to supply constraints. 10. The schedule for check-ups and the immunizations protocol vary by children’s age. 11. In a sense, the family sustainability component of Red Solidaria seems to have been put in place to deal w ith this issue, since it encompasses productive projects and microcredit schemes as income generation or diversification mechanisms. How ever, as w ill be discussed below , this component has not received much attention in the programme and its links w ith the CCT component are w eak. 12. An exception seems to be Bolsa Fam ilia in Brazil, w hich has no time limit for the receipt of benefits but uses registry updating mechanisms and consistency checks for the verification of family eligibility for the programme. 13. These additional parameters w ere not detailed in the programme’s operational manual, but in the new operational guideline for the municipal committees. They are in the process of being publicized among committee members. 14. Calculations on the basis of the statistics presented in the poverty map suggest that poverty w ould affect betw een 66 per cent and 89 per cnet of the households in the 32 municipalities of the first cluster (FLACSO, 2005). 15. In the second cluster, poverty levels w ould range betw een 52 per cent and 74 per cent (ow n calculations, based on FLACSO, 2005). 16. Although every committee receives a paper and electronic copy of the municipality’s maps, there are no specific instructions or suggestions on how they could be used and w ho should be in charge of storing and handling them. Apparently, the FISDL local representative ends up being the person responsible for them. This casts some doubt about the real use of these inputs by local stakeholders. Moreover, the mere possession of a digital format of the maps might not be sufficient to ensure that they can be utililized at the local level, as they usually require specific cartographic softw are to be view ed or printed out. 17. The operational manual of the Red includes an Indigenous Development Plan, w hich determines not only the inclusion of specific questions in the census, but also the inclusion of inter-cultural themes in the family training sessions of the CCT component of the programme and other activities that take into account specificities of the indigenous culture. For most of the programme managers interview ed, how ever, this Plan resulted much more from a donor concern than a country demand. This reflects the controversial nature of the indigenous issue in El Salvador. It is important to bear in mind that in the second half of the 20th century, El Salvador underw ent a process of denial of its indigenous ancestry. This resulted in a debate on the ‘invisibility’ of this population for the purposes of public policies (PNUD, 2003). 18. One local stakeholder mentioned the lack of possession of durable goods as one of the criteria used to validate the eligibility of urban families for the programme. 19. The impact evaluation of Red Solidaria w ill also address targeting accuracy, providing important inputs for the programme. 20. 26 of the 44 local representatives are already involved in the programme. 21. Both programmes existed before the creation of Red Solidaria. ESS w as mostly a W orld Bank funded project, w hich included the creation of mobile health units and the contracting of private health providers to reach remote communities. REE is oriented tow ards increasing educational coverage and quality by connecting schools located in the same geographical area. International Poverty Centre SBS – Ed. BNDES,10o andar 70076 900 Brasilia DF Brazil [email protected] www.undp-povertycentre.org Telephone +55 61 2105 5000 Fax +55 61 2105 5001